Hepatotoxicity has been the most common single adverse event resulting in removal of approved drugs from clinical use. Drug-induced liver injury is a leading cause of acute liver failure, and often is fatal. The overall goal of this research program is to develop a novel in vitro model to predict hepatotoxicity of drugs and derived metabolites. The mechanisms underlying drug-induced hepatotoxicity are varied and poorly understood. Recent evidence supports the hypothesis that inhibition of the hepatic canalicular bile salt export pump (Bsep), resulting in bile acid accumulation within the hepatocyte and subsequent toxicity, is an important mechanism of drug-induced liver injury. "Test Compounds" [known hepatotoxic drugs (bosentan, troglitazone, ritonavir, methotrexate, acetaminophen) and a negative control (tamoxifen)], as well as six NCI compounds, will be assayed in a novel sandwich-cultured rat hepatocyte model to test the hypothesis that Bsep inhibition by drugs/metabolites correlates with the hepatotoxic potential of these drugs. In each set of experiments, sandwich-cultured rat hepatocytes cultured in medium containing either standard or high (to maintain CYP450 activity) dexamethasone concentrations will be exposed to test compounds to assess the relative hepatotoxicity of the parent drug in the absence and presence of generated metabolites, respectively. The IC50 for drug inhibition of Bsep will be determined at concentrations insufficient to produce overt cytotoxicity during acute exposure in sandwich-cultured rat hepatocytes to avoid confounding effects of generalized cell toxicity on Bsep function. Subsequently, drug effects at the IC50 (or highest no-effect level for cytotoxicity if Bsep inhibition is not observed for a particular compound) over time in culture will be determined to define the outcome of prolonged exposure (e.g., cytotoxicity, elevated bile acid or phospholipid concentrations, altered transport protein expression). Sandwich-cultured hepatocytes represent an exciting tool for studying drug-induced hepatotoxicity because this model maintains hepatocyte polarity, morphology, liver-specific functions including metabolic activity, and allows direct access to the hepatocyte and adjacent biliary compartment. Hepatic transport proteins are expressed and localize to the correct membrane domain over time in this model. Importantly, transport function can be assessed directly in this system. Use of hepatocytes from animal species typically employed in toxicology studies (e.g., dog, monkey) as well as from humans is a particularly attractive feature of this novel in vitro model, and will be developed during the scale-up phase of this model system in a subsequent R33 application.